Item Reduction

Three initial items showed uniform differential item functioning by age. The initial 57-item analysis had properly ordered response thresholds, indicating participants used response categories as intended. The refinement process considered whether retained items represented key person-centred care domains such as information-sharing, shared decision-making, care planning, and goal-setting. Retention decisions balanced statistical evidence with theoretical considerations, conceptual value, clinical relevance, and patient-identified wording. Items were removed over five rounds of analysis. The initial full item set had psychometric problems including high fit residuals, local dependency, and failure to meet unidimensionality criteria.